Twice every year there’s discussion about getting rid of Daylight Savings Time. But this year there actually seems to be a chance that some changes will finally happen. Does anyone actually enjoy changing our clocks and adjusting our sleep schedule twice a year? Most people find the change a minor disruption but some find it to be a major disruption — especially little kids who have no problem adjusting to minor daylight changes from day to day but who will naturally stick to their prior day’s sleep schedule even if the clock has jumped forward or backward an hour since then.
We all know that Daylight Savings Time doesn’t actually “save” any daylight, and while its past purposes may not be entirely clear, there is some logic to the main intent. There are times when sunrise would be ridiculously early in some places without it and shifting the time earlier by an hour for that period can resolve those specific issues. And likewise there are times when sunset would come ridiculously early if the time were to remain shifted. So the intent is to have ideal times and shifting twice a year is what gets us that now. But is it really necessary? Is there a better way?
I’ve been thinking about these problems for years and I too have bounced around between keeping things as they are, going to no DST, or going to permanent DST. But I gave it the full analysis this morning and I think I have a good plan figured out that is sort of a 4th way to go. But first I need to run through the data that I used to study the problem.
Let’s take a look at a couple of major cities and some of the latest and earliest daylight times for 2021. [This medium post is a repost of a blog post from last year.] I’m choosing Boston and Indianapolis because they are both in the Eastern US Time Zone but have nearly an hour of daylight different between them.
First, Boston since it is further east. Let’s start with the time if Boston did not observe DST and was on Eastern Standard Time (EST) for the whole year. I’m getting all of my data from timeanddate.com.
Boston with all Eastern EST (UTC-05:00) for calendar year 2021
earliest sunrise: Jun 15, 4:06am
latest sunrise: Dec 31, 7:13am
earliest solar noon: Nov 1, 11:27am
latest solar noon: July 23, 11:50am
earliest sunset: Dec 9, 4:11pm
latest sunset: Jun 20, 7:24pm
shortest day: Dec 21, 9:05 hrs
longest day: Jun 21, 15:17 hrs
That earliest sunrise is the ridiculous one. It’s pretty early for daylight and not many people will be starting their day around then to enjoy that daylight. That’s clearly begging for a fix.
What if Boston were to stay on Daylight Savings Time year round — effectively moving Boston to the Atlantic Time Zone?
Boston with all Eastern EDT (AKA full Atlantic AST, UTC-04:00) for calendar year 2021
earliest sunrise: Jun 15, 5:06am
latest sunrise: Dec 31, 8:13am
earliest solar noon: Nov 1, 12:27pm
latest solar noon: July 23, 12:50pm
earliest sunset: Dec 9, 5:11pm
latest sunset: Jun 20, 8:24pm
shortest day: Dec 21, 9:05 hrs
longest day: Jun 21, 15:17 hrs
This doesn’t look so bad overall. That latest sunrise is a little late but sunset on that date is at 5:21 if DST were active on that date. It’s easy to see why people in New England would be in favor of being in the Atlantic Time Zone.
And now let’s look at how it comes together with DST on its current schedule.
Boston with current Eastern EST/EDT (UTC-05:00 / UTC-04:00) schedule for calendar year 2021 — DST start:Mar 14; DST end:Nov 7
earliest sunrise: Jun 15, 5:06am
latest sunrise: Dec 31, 7:13am
earliest solar noon: Nov 7, 11:27am
latest solar noon: July 23, 12:50pm
earliest sunset: Dec 9, 4:11pm
latest sunset: Jun 20, 8:24pm
shortest day: Dec 21, 9:05 hrs
longest day: Jun 21, 15:17 hrs
It’s nice that the latest sunrise is now at 7:13 instead of 8:13. But most people would consider that improvement to not be worth changing the earliest sunset to 4:11pm. So from a daylight perspective, it seems like Boston would be best served by staying on DST all year (i.e. moving to Atlantic Time).
Now let’s look at a city on the opposite edge of the Eastern Time Zone, Indianapolis. As above, let’s start with the times would be for Indianapolis if if did not observe DST. As for Boston, I’m getting my data from timeanddate.com.
Indianapolis with all Eastern EST for calendar year 2021
earliest sunrise: Jun 13, 5:15am
latest sunrise: Jan 4, 8:06am
earliest solar noon: Nov 7, 12:28pm
latest solar noon: July 25, 12:51pm
earliest sunset: Dec 7, 5:19pm
latest sunset: Jun 27, 8:16pm
shortest day: Dec 21, 9:21 hrs
longest day: Jun 21, 14:59 hrs
As you might expect for a city nearly an hour away by daylight, the hours are about an hour different from Boston’s daylight hours looking at the same time zone. And therefore while the earliest sunrise for Boston was ridiculous at 4:06am, the earliest sunrise in Indianapolis at 5:15am seems pretty reasonable. The latest sunrise is a bit late at 8:06am but when you consider that the sunset on that date would be 5:33pm, that’s not too bad — if the time were shifted to make that sunrise earlier, then the sunset would end up too early.
Now let’s look at what would happen if Indianapolis were to observe DST all year long.
Indianapolis with all Eastern EDT (AKA full Atlantic AST) for calendar year 2021
earliest sunrise: Jun 13, 6:15am
latest sunrise:J an 4, 9:06am
earliest solar noon: Nov 7, 1:28pm
latest solar noon: July 25, 1:51pm
earliest sunset: Dec 7, 6:19pm
latest sunset: Jun 27, 9:16pm
shortest day: Dec 21, 9:21 hrs
longest day: Jun 21, 14:59 hrs
Well that’s not great. Now the latest sunrise is at 9:06am. That’s now our “ridiculous” data point. And while some people may not find a 9:16pm sunset to be that weird because they are used to it, it seems pretty ridiculous to me.
And now let’s look at the times for Indianapolis working on our current DST dates.
Indianapolis with current Eastern EST/EDT schedule for calendar year 2021 — DST start: Mar 14; DST end: Nov 7
earliest sunrise: Jun 13, 6:15am
latest sunrise: Jan 4, 8:06am
earliest solar noon: Nov 7, 12:28pm
latest solar noon: July 25, 1:51pm
earliest sunset: Dec 7, 5:19pm
latest sunset: Jun 27, 9:16pm
shortest day: Dec 21, 9:21 hrs
longest day: Jun 21, 14:59 hrs
So that fixes the latest sunrise problem but that latest sunset is still pretty late. From the above data, it seems pretty clear that the best place for Indianapolis is to be on Eastern Standard time.
Looking at the data above again, note the solar noon times. That’s the point when the sun is highest in the sky. It’s not a particularly important mark on its own but when you consider that it is effectively the midpoint of the day, it becomes a bit of a shorthand for identifying when the times best match what people are are generally used to. While not everybody is fixated on the clock saying 12:00 at the time you have the solar noon, most people find that their schedules tend to be oriented around 12:00 to 12:30-ish being the midpoint of their day’s sunlight.
In Boston, the solar noon time ranges from 12:27 to 12:50 when on all DST which seems reasonable — not too far off the target. And in Indianapolis when on all DST, the solar noon time ranges from 1:28pm to 1:51pm which just seems too late to me. The better solar noon time range for Indianapolis is when on all standard Eastern time which has that range at 12:28pm to 12:51pm. When we factor in the current seasonal DST, we end up with some really wide ranges: Boston’s solar noon range is11:27am to 12:50pm and Indianapolis’s solar noon range is from 12:28pm to 1:51pm. Those wide ranges are proof to me that seasonally shifting time doesn’t really solve the underlying problem.
And what is the underlying problem? The Eastern Time zone has too many people living at the two extremes of the time zone and too few people in the sweet spot mid-point. But before I go into detail on that, I’m going to back up and let’s look at what a time zone is.
With 24 hours in a day and 360° of longitude around the world, that means that we can divide up the globe into 24 different daylight segments that are each 15° wide longitudinally. Our conterminous United States (frequently called the “contiguous United States”) starts in easternmost Maine at about 67° W longitude and goes to about 125° W in westernmost Washington state, westernmost Oregon, and westernmost California. That works out to be pretty close to 60° which is nearly exactly 4 zones of 15° each. That’s a nice simple start.
Things go off track, however, when we look at where the time zone boundaries were placed. The dividing line between Eastern and Central averages around 86°, between Central and Mountain averages around 103°, and between Mountain and Pacific averages around 115°. (I am looking at sort of the average of where the boundaries are since none of the boundaries are clean simply longitudinal lines.) That means that the Eastern time zone ends up around 19° wide, the Central time zone around 17° wide, the Mountain time zone around 12° wide and the Pacific time zone around 10°. So pretty clearly we have a width problem.
A width problem isn’t bad on its own if the width covers areas with lower populations. But in the case of our time zones, that’s not the case. We have large population centers in New York and Boston on one side of the Eastern Time zone and large population centers like Detroit, Nashville, Atlanta, and Indianapolis on the other side of the Eastern Time zone. In a way, more people in one time zone is nice when the country is thinking small and only doing business and communicating in a regional way. But with our global economy now, we are all used to dealing with people in different time locations so what may have been a benefit at one time is no longer as useful.
Let’s start over. Look at a map of the United states and now let’s group the major population centers into 4 vertical bands. There’s Boston, New York, DC, Philadelphia, Charlotte, and Raleigh-Durham in band #1, Orlando, Miami, Tampa, Cleveland, Cincinnati, Atlanta, Nashville, Indianapolis, Detroit, Chicago, St. Louis, Minneapolis, Milwaukee, and Kansas City in band #2, Houston, Dallas, Austin, Oklahoma City, Albuquerque, and Denver in band #3, and Salt Lake City, Phoenix, Las Vegas, Los Angeles, San Fransisco, San Jose, Sacramento, Portland, and Seattle in band #4. That’s where we should be aiming to have time zones and have them centered on these population bands.
(The cities in Florida are sort of on the edge of the vertical band boundary so they could end up in either bands 1 or 2 but since they are south of Georgia and since Atlanta ends up in band #2 and since the panhandle of Florida extends far west, it seemed to make sense to put the Florida cities in band #2 also.)
With this in mind we aren’t too far off with our current time zones. We just need to do a little bit of shifting of the boundaries eastward. Here’s what I propose:
After this, we might want to change the names of the new time zones. Or even better we could just stick with the current names where the blue above would be on Atlantic Time, they yellow would be Eastern, the green would be Central, and the orange would be Mountain — okay that last one would probably need to change to Western since it is both covering the former Pacific and former Mountain.
The new time zone boundaries would be closer to those ideal 15° spans and they would be better centered on the population regions. The boundary between the Atlantic and Eastern would now average around 81°, the boundary between the Eastern and Central would average around 95°, and the boundary between the Central and Western would average around 112°. The resulting widths would therefore be 14° for the Atlantic, 14° for the Eastern, 17° for the Central, and 13° for the Western. (We could get that average for the boundary between Western and Central a bit further west if we had the boundary go through Montana continuing the western border of Wyoming and that would result in a boundary average around 110° which would mean 15° widths for both Central and Western.)
After doing that, look again at my time tables above. Indianapolis would still be in the Eastern time zone which has been UTC-05:00. With no more DST, it would be UTC-05:00 year round which keeps it with the times listed for Eastern Standard time which the data above showed to be the best for it. Boston would move to the Atlantic Time Zone which is UTC-04:00 — or the same as it is now when DST is active — except it would be yearround. And as described above, those are the times that seem to make the most sense for Boston. And to achieve these more ideal times, no seasonal time shift is needed and the solar noon remains more or less in the middle of the day.
